def _load_default_texts():
"""
Loads default general texts
Returns
-------
result : default 20newsgroup texts
"""
dataset = Dataset()
dataset.fetch_dataset("20NewsGroup")
return dataset.get_corpus()
def test_model_output_lda_tomotopy(data_dir):
dataset = Dataset()
dataset.load_custom_dataset_from_folder(data_dir + '/M10')
num_topics = 3
model = LDATOMOTO(num_topics=num_topics, alpha=0.1)
output = model.train_model(dataset)
assert 'topics' in output.keys()
assert 'topic-word-matrix' in output.keys()
assert 'test-topic-document-matrix' in output.keys()
# check topics format
assert type(output['topics']) == list
assert len(output['topics']) == num_topics
# check topic-word-matrix format
assert type(output['topic-word-matrix']) == np.ndarray
assert output['topic-word-matrix'].shape == (num_topics,
len(dataset.get_vocabulary()))
# check topic-document-matrix format
assert type(output['topic-document-matrix']) == np.ndarray
assert output['topic-document-matrix'].shape == (
num_topics, len(dataset.get_partitioned_corpus()[0]))
# check test-topic-document-matrix format
assert type(output['test-topic-document-matrix']) == np.ndarray
assert output['test-topic-document-matrix'].shape == (
num_topics, len(dataset.get_partitioned_corpus()[2]))
def test_model_output_ctm_combined(data_dir):
dataset = Dataset()
dataset.load_custom_dataset_from_folder(data_dir + '/M10')
num_topics = 3
model = CTM(num_topics=num_topics, num_epochs=5, inference_type='combined')
output = model.train_model(dataset)
assert 'topics' in output.keys()
assert 'topic-word-matrix' in output.keys()
assert 'test-topic-document-matrix' in output.keys()
# check topics format
assert type(output['topics']) == list
assert len(output['topics']) == num_topics
# check topic-word-matrix format
assert type(output['topic-word-matrix']) == np.ndarray
assert output['topic-word-matrix'].shape == (num_topics,
len(dataset.get_vocabulary()))
# check topic-document-matrix format
assert type(output['topic-document-matrix']) == np.ndarray
assert output['topic-document-matrix'].shape == (
num_topics, len(dataset.get_partitioned_corpus()[0]))
# check test-topic-document-matrix format
assert type(output['test-topic-document-matrix']) == np.ndarray
assert output['test-topic-document-matrix'].shape == (
num_topics, len(dataset.get_partitioned_corpus()[2]))
def test_model_output_nmf(data_dir):
dataset = Dataset()
dataset.load_custom_dataset_from_folder(data_dir + '/M10')
num_topics = 3
model = NMF(num_topics=num_topics,
w_max_iter=10,
h_max_iter=10,
use_partitions=True)
output = model.train_model(dataset)
assert 'topics' in output.keys()
assert 'topic-word-matrix' in output.keys()
assert 'test-topic-document-matrix' in output.keys()
# check topics format
assert type(output['topics']) == list
assert len(output['topics']) == num_topics
# check topic-word-matrix format
assert type(output['topic-word-matrix']) == np.ndarray
assert output['topic-word-matrix'].shape == (num_topics,
len(dataset.get_vocabulary()))
# check topic-document-matrix format
assert type(output['topic-document-matrix']) == np.ndarray
assert output['topic-document-matrix'].shape == (
num_topics, len(dataset.get_partitioned_corpus()[0]))
# check test-topic-document-matrix format
assert type(output['test-topic-document-matrix']) == np.ndarray
assert output['test-topic-document-matrix'].shape == (
num_topics, len(dataset.get_partitioned_corpus()[2]))
def test_model_output_prodlda_not_partitioned(data_dir):
dataset = Dataset()
dataset.load_custom_dataset_from_folder(data_dir + '/M10')
num_topics = 3
model = ProdLDA(num_topics=num_topics, num_epochs=5, use_partitions=False)
output = model.train_model(dataset)
assert 'topics' in output.keys()
assert 'topic-word-matrix' in output.keys()
assert 'test-topic-document-matrix' not in output.keys()
# check topics format
assert type(output['topics']) == list
assert len(output['topics']) == num_topics
# check topic-word-matrix format
assert type(output['topic-word-matrix']) == np.ndarray
assert output['topic-word-matrix'].shape == (num_topics, len(dataset.get_vocabulary()))
# check topic-document-matrix format
assert type(output['topic-document-matrix']) == np.ndarray
assert output['topic-document-matrix'].shape == (num_topics, len(dataset.get_corpus()))
def test_partitions_custom(data_dir):
dataset = Dataset()
dataset.load_custom_dataset_from_folder(data_dir + "M10")
partitions = dataset.get_partitioned_corpus()
assert len(partitions[0]) == 5847
assert len(partitions[1]) == 1254
def test_partitions_fetch():
dataset = Dataset()
dataset.fetch_dataset("M10")
partitions = dataset.get_partitioned_corpus()
assert len(partitions[0]) == 5847
assert len(partitions[1]) == 1254
def test_load_M10():
dataset = Dataset()
dataset.fetch_dataset("M10")
assert len(set(dataset.get_labels())) == 10
def test_load_20ng():
data_home = get_data_home(data_home=None)
cache_path = _pkl_filepath(data_home, "20NewsGroup" + ".pkz")
if os.path.exists(cache_path):
os.remove(cache_path)
dataset = Dataset()
dataset.fetch_dataset("20NewsGroup")
assert len(dataset.get_corpus()) == 16309
assert len(dataset.get_labels()) == 16309
assert os.path.exists(cache_path)
dataset = Dataset()
dataset.fetch_dataset("20NewsGroup")
assert len(dataset.get_corpus()) == 16309
def preprocess_dataset(self, documents_path, labels_path=None):
"""
preprocess the input dataset
:param documents_path: path to the documents file. Each row of the file represents a document
:type documents_path: str
:param labels_path: path to the documents file. Each row of the file represents a label. Its index corresponds
to the index of the documents file (default: None)
:type labels_path: str
:return octis.dataset.dataset.Dataset
"""
docs = [line.strip() for line in open(documents_path, 'r').readlines()]
if self.num_processes is not None:
# with Pool(self.num_processes) as p:
# docs = p.map(self.simple_preprocessing_steps, docs)
docs = process_map(self.simple_preprocessing_steps, docs, max_workers=self.num_processes, chunksize=1)
else:
docs = self.simple_preprocessing_steps(docs)
if self.lowercase:
self.preprocessing_steps.append("lowercase")
if self.remove_punctuation:
self.preprocessing_steps.append('remove_punctuation')
if self.lemmatize:
self.preprocessing_steps.append('lemmatize')
vocabulary = self.filter_words(docs)
print("created vocab")
# with Pool(self.num_processes) as p:
# final_docs, final_labels = p.starmap(self._foo, product(docs, vocabulary, labels_path, repeat=2))
print(len(vocabulary))
final_docs, final_labels, document_indexes = [], [], []
if labels_path is not None:
labels = [line.strip() for line in open(labels_path, 'r').readlines()]
for i, doc, label in zip(range(len(docs)), docs, labels):
vocab = set(vocabulary)
new_doc = [w for w in doc.split() if w in vocab]
if len(new_doc) > self.min_doc_words:
final_docs.append(new_doc)
final_labels.append(label)
document_indexes.append(i)
else:
for i, doc in enumerate(docs):
vocab = set(vocabulary)
new_doc = [w for w in doc.split() if w in vocab]
if len(new_doc) > self.min_doc_words:
final_docs.append(new_doc)
document_indexes.append(i)
self.preprocessing_steps.append('filter documents with less than ' + str(self.min_doc_words) + " words")
if self.verbose:
print("words filtering done")
metadata = {"total_documents": len(docs), "vocabulary_length": len(vocabulary),
"preprocessing-info": self.preprocessing_steps, "labels": list(set(final_labels)),
"total_labels": len(set(final_labels))}
if self.split:
if len(final_labels) > 0:
train, test, y_train, y_test = train_test_split(
range(len(final_docs)), final_labels, test_size=0.15, random_state=1, stratify=final_labels)
train, validation = train_test_split(train, test_size=3 / 17, random_state=1, stratify=y_train)
partitioned_labels = [final_labels[doc] for doc in train + validation + test]
partitioned_corpus = [final_docs[doc] for doc in train + validation + test]
document_indexes = [document_indexes[doc] for doc in train + validation + test]
metadata["last-training-doc"] = len(train)
metadata["last-validation-doc"] = len(validation) + len(train)
if self.save_original_indexes:
return Dataset(partitioned_corpus, vocabulary=vocabulary, metadata=metadata, labels=partitioned_labels,
document_indexes=document_indexes)
else:
return Dataset(partitioned_corpus, vocabulary=vocabulary, metadata=metadata, labels=partitioned_labels)
else:
train, test = train_test_split(range(len(final_docs)), test_size=0.15, random_state=1)
train, validation = train_test_split(train, test_size=3 / 17, random_state=1)
metadata["last-training-doc"] = len(train)
metadata["last-validation-doc"] = len(validation) + len(train)
partitioned_corpus = [final_docs[doc] for doc in train + validation + test]
document_indexes = [document_indexes[doc] for doc in train + validation + test]
if self.save_original_indexes:
return Dataset(partitioned_corpus, vocabulary=vocabulary, metadata=metadata, labels=final_labels,
document_indexes=document_indexes)
else:
return Dataset(partitioned_corpus, vocabulary=vocabulary, metadata=metadata, labels=final_labels,
document_indexes=document_indexes)
else:
if self.save_original_indexes:
Dataset(final_docs, vocabulary=vocabulary, metadata=metadata, labels=final_labels,
document_indexes=document_indexes)
else:
Dataset(final_docs, vocabulary=vocabulary, metadata=metadata, labels=final_labels)
def dataset(data_dir):
# Load dataset
dataset = Dataset()
dataset.load_custom_dataset_from_folder(data_dir + '/M10')
return dataset
def dataset(root_dir):
dataset = Dataset()
dataset.load_custom_dataset_from_folder(root_dir + "/../preprocessed_datasets/" + '/M10')
return dataset
def _restore_parameters(self, name_path):
"""
Restore the BO parameters from the json file
:param name_path: name of the json file
:type name_path: str
:return: result of BO optimization (scikit-optimize object), surrogate model (scikit-learn object)
:rtype: tuple
"""
# Load the previous results
with open(name_path, 'rb') as file:
optimization_object = json.load(file)
self.search_space = load_search_space(optimization_object["search_space"])
self.acq_func = optimization_object["acq_func"]
self.surrogate_model = optimization_object["surrogate_model"]
self.kernel = eval(optimization_object["kernel"])
self.optimization_type = optimization_object["optimization_type"]
self.model_runs = optimization_object["model_runs"]
self.save_models = optimization_object["save_models"]
self.save_step = optimization_object["save_step"]
self.save_name = optimization_object["save_name"]
self.save_models = optimization_object["save_models"]
self.save_path = optimization_object["save_path"]
self.early_stop = optimization_object["early_stop"]
self.early_step = optimization_object["early_step"]
self.plot_model = optimization_object["plot_model"]
self.plot_best_seen = optimization_object["plot_best_seen"]
self.plot_name = optimization_object["plot_name"]
self.log_scale_plot = optimization_object["log_scale_plot"]
self.random_state = optimization_object["random_state"]
self.dict_model_runs = optimization_object['dict_model_runs']
self.number_of_previous_calls = optimization_object['current_call'] + 1
self.current_call = optimization_object['current_call'] + 1
self.number_of_call = optimization_object['number_of_call']
self.save_path = optimization_object['save_path']
self.x0 = optimization_object['x0']
self.y0 = optimization_object['y0']
self.n_random_starts = optimization_object['n_random_starts']
self.initial_point_generator = optimization_object['initial_point_generator']
self.topk = optimization_object['topk']
self.time_eval = optimization_object["time_eval"]
res = None
# Load the dataset
dataset = Dataset()
if not optimization_object["is_cached"]:
dataset.load_custom_dataset_from_folder(optimization_object["dataset_path"])
else:
dp = optimization_object["dataset_path"][:-(len(optimization_object["dataset_name"]) + len("_py3.pkz"))]
dataset.fetch_dataset(optimization_object["dataset_name"], data_home=dp)
self.dataset = dataset
# Load the metric
self._load_metric(optimization_object, dataset)
# Load the model
self.model = load_model(optimization_object)
# Creation of the hyperparameters
self.hyperparameters = list(sorted(self.search_space.keys()))
# Choice of the optimizer
opt = choose_optimizer(self)
# Update number_of_call for restarting
for i in range(self.number_of_previous_calls):
next_x = [optimization_object["x_iters"][key][i] for key in self.hyperparameters]
f_val = -optimization_object["f_val"][i] if self.optimization_type == 'Maximize' else \
optimization_object["f_val"][i]
res = opt.tell(next_x, f_val)
# Create the directory where the results are saved
Path(self.save_path).mkdir(parents=True, exist_ok=True)
self.model_path_models = self.save_path + "models/"
return res, opt